ID: 5d8a0da9ef1a6d55dbca34c1
Head pose estimation
by SHIVAM GARG
Calculating pitch yaw roll using deep learning
License: MIT license
Tags:
Model stats and performance
Framework | Tensorflow |
OS Used | Linux |
Inference time in seconds per sample.
Screenshots
HEADPOSE ESTIMATION USING FACIAL LANDMARKS AND OPENCV
WHAT IS IT?
There are three major steps: Face detection: A face detector is adopted to provide a face box containing a human face. Then the face box is expanded and transformed to a square to suit the needs of later steps. Facial landmark detection: A custom trained facial landmark detector based on TensorFlow is responsible for output 68 facial landmarks. Pose estimation: Once we got the 68 facial landmarks, a mutual PnP algorithms is adopted to calculate the pose. The marks is detected frame by frame, which result in small variance between adjacent frames. This makes the pose unstable. A Kalman filter is used to solve this problem, one can draw the original pose to observe the difference.
HOW TO USE?
To run the script
python run.py -input test.jpg -gpu_frac 0
For help options - python run.py -h
usage:
run.py [-h] [-input INPUT] [-gpu_frac GPU_FRAC]
optional arguments: -h, --help show this help message and exit -input INPUT mention input image -gpu_frac GPU_FRAC mention gpu fraction to use
WHAT ARE THE REQUIREMENTS?
To get all the requirements and dependencies installed run the command
For GPU - pip install -r gpu_requirements.txt
For CPU - pip install -r cpu_requirements.txt
Author View Profile
Views
A philosophy student cleverly disguised as a Coax Deep Learning engineer spending whole day, practically every day, experimenting with TensorFlow,Pytorch, and Caffe; dabbling with Python and C++; and drinking a wide variety of Coffee everyday.
User Reviews
0 total ratings
Model has not been reviewed yet.